I started this blog before Facebook, Instagram and Twitter were founded. In those days, if people wanted to find out what you were up to they went to your blog. I had a small readership of friends plus some acquaintances from the hacking scene.
Social media getting everyone to go to a single site/app for their updates and my changing interests decimated my readership. I don’t know how many regulars are on here anymore, maybe two.
I’ve seen this blog cited in papers, news articles, social media posts, and a book – so I know that I have a bit more irregular traffic. But my point is that the days of this being a little blog with other little blogs out there, kind of like how people digest zines now, are over. This is a relic of a time that is no more.
And that’s okay. I still write.
I do it because for all the self-censorship I still do, it’s as close to an archive of my life as there will ever be. I’ve run this site since I was a teenager, and my life literally flashes before my eyes on here.
I continue to write not for the benefit of others but my own.
Enter AI
Every word on his site has been shaped by who I am and how I see the world. My essence is not on here but its fingerprints are.
Enter AI and specifically large language models (LLMs). These have taken over the world since going mainstream in 2022. Four years later, and now half of new articles on the Internet are AI generated. At my company, the CEO used AI to compose a “heartfelt” email about the recent mass lay-offs, but it’s used everywhere else internally as well.
Large language models are word prediction machines. A person provides a prompt and it uses statistics derived from the world’s written content to come up with what words are likely to follow in a response.
It’s an impressive technical achievement. But it’s incapable of two things: emotion and art.
Emotion because the selected words are merely the statistical distillation of books and whatever else these AI companies fed its models. There is no genuine human connection or feeling.
That’s why the CEO’s email was so problematic. In a moment that required a bit of humanity, where staff felt disposable, he delegated the task of appearing empathetic to a machine incapable of it.
In much the same way, AI is incapable of making art. It can produce a convincing facsimile. But does not experience anything new and is incapable of original thought. The ingredients of humanity’s greatest artists are absent. AI then is merely a derivative works machine. Art exists to communicate something about ourselves but AI can only emulate its aesthetics.
Much like the CEO’s plagiarized email could only emulate the aesthetics of being heartfelt. The humanity was in the one thing he did not communicate: the prompt he supplied the LLM.
Back to the Blog
The reason I’ll never use AI – in the 2026 sense of the word – on this blog is because it would defeat the purpose of its existence: to share the way that I am different with you. An LLM would dilute that. There is nothing that it could add that would be worth reading.
I create what I wanted to read myself, and I don’t want to read (or watch or listen to) AI slop.
I think there is a place for this automation in other domains such as finding cancers, translating for tourists, helping people code, car safety mechanisms, etc.
But it has no heart and therefore no place by mine.